Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH net-next 2/2] bridge: implement get_link_ksettings ethtool
 method
To:     Matthias Schiffer <mschiffer@universe-factory.net>,
        netdev@vger.kernel.org
Cc:     linux-kernel@vger.kernel.org, davem@davemloft.net,
        roopa@cumulusnetworks.com
References: <cover.1573321597.git.mschiffer@universe-factory.net>
 <8e414e98928aba7f11ea997498fb18af05434f5f.1573321597.git.mschiffer@universe-factory.net>
From:   Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Message-ID: <5da07a98-bae3-2d67-d60b-2e58be9d8377@cumulusnetworks.com>
Date:   Tue, 12 Nov 2019 00:15:24 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.1.1
MIME-Version: 1.0
In-Reply-To: <8e414e98928aba7f11ea997498fb18af05434f5f.1573321597.git.mschiffer@universe-factory.net>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 09/11/2019 19:54, Matthias Schiffer wrote:
> We return the maximum speed of all active ports. This matches how the link
> speed would give an upper limit for traffic to/from any single peer if the
> bridge were replaced with a hardware switch.
> 
> Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
> ---
>  net/bridge/br_device.c | 37 +++++++++++++++++++++++++++++++++++--
>  1 file changed, 35 insertions(+), 2 deletions(-)
> 
> diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c
> index e804a3016902..e8a2d0c13592 100644
> --- a/net/bridge/br_device.c
> +++ b/net/bridge/br_device.c
> @@ -263,6 +263,38 @@ static void br_getinfo(struct net_device *dev, struct ethtool_drvinfo *info)
>  	strlcpy(info->bus_info, "N/A", sizeof(info->bus_info));
>  }
>  
> +static int br_get_link_ksettings(struct net_device *dev,
> +				 struct ethtool_link_ksettings *cmd)
> +{
> +	struct net_bridge *br = netdev_priv(dev);
> +	struct net_bridge_port *p;
> +
> +	cmd->base.duplex = DUPLEX_UNKNOWN;
> +	cmd->base.port = PORT_OTHER;
> +	cmd->base.speed = SPEED_UNKNOWN;
> +
> +	list_for_each_entry(p, &br->port_list, list) {
> +		struct ethtool_link_ksettings ecmd;
> +		struct net_device *pdev = p->dev;
> +
> +		if (!netif_running(pdev) || !netif_oper_up(pdev))
> +			continue;
> +
> +		if (__ethtool_get_link_ksettings(pdev, &ecmd))
> +			continue;
> +
> +		if (ecmd.base.speed == (__u32)SPEED_UNKNOWN)
> +			continue;
> +
> +		if (cmd->base.speed == (__u32)SPEED_UNKNOWN ||
> +		    cmd->base.speed < ecmd.base.speed) {
> +			cmd->base.speed = ecmd.base.speed;
> +		}
> +	}
> +
> +	return 0;
> +}
> +
>  static netdev_features_t br_fix_features(struct net_device *dev,
>  	netdev_features_t features)
>  {
> @@ -365,8 +397,9 @@ static int br_del_slave(struct net_device *dev, struct net_device *slave_dev)
>  }
>  
>  static const struct ethtool_ops br_ethtool_ops = {
> -	.get_drvinfo    = br_getinfo,
> -	.get_link	= ethtool_op_get_link,
> +	.get_drvinfo		 = br_getinfo,
> +	.get_link		 = ethtool_op_get_link,
> +	.get_link_ksettings	 = br_get_link_ksettings,
>  };
>  
>  static const struct net_device_ops br_netdev_ops = {
> 

Code-wise this is ok, but it could cause potential issues because devices will suddenly
have speed (without duplex), and macvlan/8021q are examples which use the lowerdev's
get_link_ksettings. I searched for potential problems with some user-space software
but couldn't find any, so I hope it will be fine. :)
Let's see where this goes.

Acked-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>